EurJ f,nmr Clrn OncoL Vol. 24, No I I, pp. 1715-1719, 1988. Printed in Grear Brirain

Comparison of Enzyme Immunoassay with Dextran-coated Charcoal Method in the Determination of Progesterone Receptor in Breast Cancer Cytosols

SHINZABURO NOGUCHI, KEISUKE MIYAUCHI, SHINGI IMAOKA, HIROKI KOYAMA and TAKESHI IWANAGA

Department of Surgery, The Center for Adult Diseases, Osaka, 3-Nakamichi l-Chome, Higashinari-ku, Osaka 537, Japan

Abstract-A new enzyme immunoassay (EZA) for progesterone receptor (PR) has been developed by Abbott Laboratories. To study the correlation of the results from this new technique with the currently existing tritiated-ligand binding assay [dextran-coated charcoal (DCC) method], cylosols from 70 human breast caruers were assayed for PR by both the EZA and DCC methods. EZA showed a good reproducibilip and was not susceptible to the change in protein concentration of cytosols and #es of reducing agents. The correlation between the EZA and DCC methods was excellent with a correlation coef&ent of 0.946; regression curve was PR (EZA) = 1.06 X PR(DCC) + 0.19 fmolfmg protein. The concordance of the results obtained

from both methods was 91.4% with a cutoff value of 10 fmollmg protein. These results demonstrate that EZA is a very useful and reliable method in the determination of PR with an excellent correlation with the conventional DCC method.

INTRODUCTION IT IS NOW accepted that measurement ofthe estrogen receptor (ER) and progesterone receptor (PR) in breast cancer is clinically useful for predicting the response of patients to endocrine therapy and for determining their prognosis [ 1, 21. These diagnostic assays seem essential for planning the treatment of breast cancer patients.

Several methods have been developed for the determination of ER and PR in breast cancer and all of them are based on the evaluation of the [3H]steroid binding capacity of breast cancer cyto- ~01s. These binding assays have been the main methods used during the past 20 years and the clinical significance of ER and PR have been evalu- ated on the results obtained from these assays. [3H]Ligand b’ d’ g m m assays, however, are not devoid of problems. Since these assays can only detect the receptors in the unoccupied form, interference by endogenous hormones cannot be eliminated com- pletely. Another limitation of these assays is that small samples cannot be assayed since a multipoint binding assay usually requires at least 200-300 mg of tissue. Inclusion of tritiated ligands in the assay procedure is another demerit and is not suited for the routine laboratory assay.

Accepted 22 June 1988. Author to whom correspondence and reprint requests should be sent: Shinzaburo Noguchi M.D., Department of Surgery, The Center for Adult Diseases, 3-Nakamichi I-Chome, Higashinari- ku, Osaka 537, Japan.

&Ic 24:11-c 1715

Recently, an ER enzyme immunoassay (ER-EIA) kit has been introduced and an excellent correlation between the dextran-coated charcoal (DCC) method and ER-EIA has been demonstrated [3,4]. The ER-EIA kit presently available is based on the direct recognition of the ER molecule by a monoclonal antibody raised against ER purified from MCF-7 cells. The merits of ER-EIA are as follows: (1) ER-EIA can detect ER whether or not it is occupied by endogenous estrogens, this advantage has enabled the application of this method to the exchange assay of ER [5]; (2) only a small sample volume is required for the assay and even specimens obtained from fine needle biopsy or drill biopsy can be assayed accurately with this method [6]; (3) exclusion of tritiated ligands from the assay procedure makes ER-EIA more suited for routine laboratory examination.

More recently, Greene et al. have succeeded in producing monoclonal antibody against PR purified from T47D cells [7,8] and Abbott Laboratories has produced a PR-EIA kit based on this monoclonal antibody, which was demonstrated to recognize both the 95 K and 120 K forms of PR. If PR-EIA can detect PR in cytosols from human breast cancer with accuracy, it will confer the same advantages over conventional DCC method as described for ER-EIA. However, this point has not been studied sufficiently.

In this paper, a comparison of PR-EIA with the conventional DCC method was made on cytosols from 70 human breast cancers. The reliability of

1716 S. Noguchi et al.

the PR-EIA kit and influence of various assay conditions on PR-EIA results were also studied.

MATERIAIS AND METHODS

Preparation of cytosols from surgical specimens Seventy patients with operable breast cancer were

included in this study. Surgically obtained breast cancer specimens were stripped of blood and necrotic tissue and snap frozen in a deep freezer at -80°C. These specimens were kept in this condition until the assay, which was performed in no more than 3 months after the surgery.

All procedures were carried out at 0 - 4°C unless noted otherwise. The specimens were minced with scissors and homogenized in seven volumes of homo- genization buffer [lo mM Tris, 1.5 mM EDTA, 5 mM monothioglycerol, 10 mM sodium molyb- date, 10% (v/v) glycerol, pH 7.41 using a Polytron PT homogenizer (Brinkmann Instruments, Inc., Westbury, NY) set at 4, with three 10-s runs allowing 30 s for cooling between each run. The homogenate was centrifuged at 105,000 g for 60 min and supernatant cytosol was obtained with- out the superficial lipid layer. Cytosols were diluted with homogenization buffer so that they contained a protein concentration of l-2 mg/ml.

Tritiated-ligand binding assay for PR (dextran-coated charcoal method)

The cytosol was incubated with 0.1-10 nM (1 7a-methyl-[3H])-p romegestone (86.3 Ci/mmol, New England Nuclear, Boston, MA) in the presence and absence of a 250-fold molar excess of cold promegestone for 18 h (reaction volume, 400 ~1). After the incubation, 400 pl of dextran-coated char- coal solution (10 mM Tris, 0.5% Norit A, 0.005% Dextran T-70, pH 8.0) was added and incubated for 30 min with intermittent vortexing. Then, the mixture was centrifuged at 3000 rpm for 10 min and 200 p,l of the supernatant was assayed for radioactivity in a liquid scintillation counter. The maximum binding sites and Kd were estimated by Scatchard plot analyses [9].

Enzyme immunoassay for ER and PR The enzyme immunoassay kit for ER (ER-EIA)

was purchased from Abbott Laboratories, Chicago, IL. ER-EIA is a solid phase enzyme immunoassay based on the ‘sandwich’ principle. All the pro- cedures for the ER-EIA assay were performed according to the manufacturer’s instructions. In brief, cytosols, controls and standards were incu- bated with beads coated with anti-ER monoclonal antibody (rat) for 18 h at 4°C and unbound material was removed by aspiration and washing ofthe beads with distilled water. A second anti-ER monoclonal antibody (rat) conjugated horseradish peroxidase

was incubated with the ER-bead complex at 37°C for 1 h. After aspiration of excess conjugate and washing of the beads with distilled water, the beads were incubated in enzyme substrate (hydrogen per- oxide and o-phenylenediamine) for 30 min at room temperature. The reaction was stopped by the addition of 1 N sulfuric acid. The intensity of color developed was recorded at 492 nm using a Quan- tum II. A standard curve was obtained by plotting the absorbance ofthe standards vs. their ER concen- trations, and ER concentrations of samples were determined from this standard curve.

Enzyme immunoassay kit for PR (PR-EIA) was a generous gift from Dainabot Co. Ltd. (Tokyo, Japan). All the assay procedures were performed according to the manufacturer’s instructions. The principle on which PR-EIA is based is identical to that of ER-EIA and the assay procedure of PR-EIA is quite similar to that of ER-EIA except for only the difference in temperature (4°C) at the second incubation.

Miscellaneous assays The protein concentrations were assayed accord-

ing to the method of Lowry et al. [IO].

RESULTS

Reproducibility of ER-EIA and PR-EIA Assay reproducibility was determined by assay-

ing one control (included in each kit) in quintupli- cate in the same run (intra-assay variation) and in duplicate in five independent runs (inter-assay variation). The coefficient of variation (% CV) was determined from the components of variance. The excellent reproducibility of ER-EIA has already been reported by many investigators and our results (Table 1) are compatible with those reported by others. % CV of PR-EIA (intra-assay variation, 5.6%; inter-assay variation, 8.7%) were as low as that of ER-EIA (intra-assay variation, 4.1%) inter- assay variation, 7.9%). Therefore, the PR-EIA kit was considered to have an excellent reproducibility identical to that of ER-EIA.

Injuence of protein concentration on PR-EIA The influence of protein concentration of cytosols

on the assay results of PR-EIA was studied (Fig. 1). Each cytosol from 26 breast cancers was diluted three-fold with homogenization buffer, and assayed both in the diluted and non-diluted forms by PR- EIA. The mean value of each cytosol (fmol/mg protein) was taken as 100% and the percentage deviation from the mean was plotted in relation to protein concentration (0.5-6.0 mg/ml). A good stability of assay results was observed over this range of protein concentrations.

Enzyme Immunoassay for PR 1717

Table 1. Intra- and inter-assay variation of ER-EIA and PR- EIA

Intra-assay variation* Mean S.D. % CV

Inter-assay variation7 Mean S.D. %CV

ER-EIA PR-EIA

118.2 57.4 4.8 3.2 4.1 5.6

122.4 55.2 9.6 4.8 7.9 a.7

*Intra-assay variation was determined by assaying one control in quintuplicate in the same run. tinter-assay variation was determined by assaying one control in duplicate in five independent runs.

%

OJ I , I

0 1 2 3 4 5 6

Protein Concwftration (mg/mP)

Fig. 1. Each @sol from 26 breast cancers was diluted with homogeniz- ation buffer three-fold and was assayed in the diluted (a) and non-diluted (0) forms by PR-EIA. The mean result of each cytosol was taken as 100% and thepercentage deviationfrom the mean wasplotted in relation to

protein concentration.

InJuence of reducing agents on PR-EIA Reducing agents are usually used in the homogen-

ization buffer in conventional tritiated-ligand bind- ing assays in order to stabilize and protect receptor molecules. As reducing agents, monothioglycerol and dithiothreitol are most commonly used. We studied the influence of the reducing agents on PR- EIA (Fig. 2). In this experiment, ten tumors were homogenized in seven volumes of reducing-agent- free buffer [lo mM Tris, 1.5 mM EDTA, 10 mM sodium molybdate, 10% (v/v) glycerol, pH 7.41 and supernatant cytosols were obtained after the centrifugation at 105,000 g for 60 min. Each cyto- sol was diluted two-fold with the homogenization buffer containing 10 mM monothioglycerol or 1 mM dithiothreitol and then assayed for PR by PR- EIA. The correlation between PR content obtained from monothioglycerol-containing buffer and those obtained from dithiothreitol-containing buffer is illustrated in Fig. 2. The results obtained under both conditions were almost identical (relative coef- ficient was 0.991 and regression curve was

Y=O.99X+2.5

r =0.991

“oy, , , 50 100 150

PR Values (fmd/mcp) in DTT Buffer

Fig. 2. Ten cytosols were assayed @ PR-EIA in both monothiog!ycerol (MTG, 5 mM) containing buffer and dithiothreitol (DTT, 0.5 mM) containing buffer. Linear regression analysis gave the following results: correlation coefficient (r) = 0.991 and regression curve; y = 0.991x +

2.5. mcp; mg cytosol protein.

y = 0.99x + 2.5). Therefore, a difference in the reducing agents did not seem to influence PR-EIA.

InJuence of receptor occupancy by hormones on EIA The influence of estrogen binding to ER and

progesterone binding to PR on EIA results was studied by incubating five cytosols with 10 nM 17@estradiol or 10 nM promegestone for 2 h before the ER-EIA or PR-EIA (Table 2). ER contents obtained in the presence and absence of cold 17p- estradiol were almost identical and PR contents obtained in the presence and absence of promege- stone were also almost identical. Therefore, PR- EIA, like ER-EIA, is considered to detect PR mol- ecules independent of endogenous progesterone binding.

Correlation between the DCC method and PR-EIA Cytosols from 70 breast cancers were assayed for

PR by both the DCC method and the PR-EIA. The correlation of PR contents obtained from both methods is illustrated in Fig. 3. An excellent corre- lation was found between PR-EIA and DCC method (r = 0.946) and regression curve was Y = 1.06X + 0.19 (Y, PR-EIA; X, DCC method).

Regression curves calculated according to the menopausal status were as follows; Y = 1.13X + 4.53 (r = 0.948) for 31 premenopausals and Y = 0.99X - 1 .I3 (r = 0.957) for 39 postmeno- pausals. No significant difference was found in the slope of regression curves between premenopausals and postmenopausals.

DISCUSSION Our results demonstrate that the PR-EIA kit has

1718 S. Noguchi et al.

Table 2. ER and PR values assayed by EIA before and after treatment with estradiotand R5020, re$ectively

Treatment with R5020

Before After

Treatment with estradiol

Before After Sample

1 108* 95 103 90 2 95 92 66 87 3 48 51 26 28 4 II 8 13 12 5 0 0 0 0

*Mean (fmol/mg protein) oftriplicatedeterminations.

300

1

v=1.06x+0.19

/

.

r SO.946 .

.

z 200-

F u E a

! % loo-

.’ 0 I 0 1

0 loo 200 300

DCC Method (fmd/mcp)

Fig. 3. Cytosols from 70 breast cancers were assayedfor PR by both the PR-HA and DCC methods. Correlation of PR values obtainedfrom both methods is illustrated. Linear regression anabsis gave the result of

correlation coe$icicnt (r) = 0.946 and regression curve; y = 1.06.x + 0.19. mcp; mg cytosol protein.

good reproducibility and is not susceptible to change in protein concentration of cytosols and types of reducing agents routinely used in the homogeniz- ation buffer and that the correlation between tritia- ted-ligand binding assay and EIA for progesterone receptor in breast cancer cytosols is excellent with a correlation coefficient of 0.946; the slope of the regression curve is 1.06 with the Y-axis intercept not significantly different from zero. Therefore, PR- EIA appears to be a very useful method for PR assay with advantages over the conventional DCC method because of the small sample volume required for the assay and the elimination of the need to use radioisotopes in the assay procedure.

Another important advantage of EIA lies in its ability to detect both occupied and unoccupied receptors, in contrast to the DCC method which can measure only unoccupied receptors. With tumors from premenopausal patients, some receptors are thought to be occupied by endogenous hormones, and in such cases the DCC method seems to under- estimate the level of receptors while EIA can detect

the precise level. Goussard et al. reported that ER levels of tumors measured by ER-EIA was significantly higher than those measured by the DCC method in premenopausal patients though ER levels measured by ER-EIA were close to those measured by the DCC method in postmenopausal patients [3]. However, the present results concern- ing PR assay suggest that endogenous progesterones do not seem to play an important role since no significant ‘difference was found in PR levels meas- ured by both methods irrespective of menopausal status. This result is partly attributable to the difference in exchangeability between ER and PR. Estrogen bound to ER exchanges with [3H]estradiol at 3O”C, but not at O-4% where the DCC method for ER assay is usually performed [ 111. Therefore, when there are some occupied ERs, the DCC method may underestimate the ER value since it can only detect unoccupied ERs. On the other hand, PR is known to release progesterone more easily than ER does estrogen even at O-4%. Bayard et al. demonstrated that progesterone bound to PR easily exchanges with [3H]progesterone at O-4% [12]. Even in the presence of occupied PRs, the DCC method is considered to detect both occupied and unoccupied PRs. Therefore, interference of endogenous progesterones are less likely to occur in the PR assay as compared to the ER assay.

ER-EIA has been demonstrated to be well suited for the determination of nuclear ER (ERn) in high salt extracts ofmyofibrillar pellets [5]. Since conven- tional exchange assays for ERn are sensitive to procedural details and have unavoidable problems such as ER degradation during assay and relatively large amounts of tissue required for the assay, ER- EIA, which is almost devoid of the above-mentioned problems, will be a very useful tool for an accurate determination of ERn. Probably, PR-EIA can also be used for a PRn assay because PR-EIA can detect both unoccupied and occupied forms of PR and high salt conditions, which are necessary for the extraction of PRn from myofibrillar pellets, do not affect PR-EIA (data not shown).

DCC methods currently in use in laboratories vary from one to another in procedural details such as composition of buffer and incubation conditions. This situation is inconvenient for the interlaboratory comparison of PR data. However, PR-EIA will solve this problem due to a total standardization of assay procedures. Moreover, a correlation line between the PR-EIA and DCC methods obtained in our series was approximately 1.0 with a Y- axis intercept not significantly different from zero. Therefore, serious clinical problems are unlikely to occur in the interpretation of PR data after the passage of the DCC method to PR-EIA. Due to the excellent reproducibility and the other advantages mentioned above, PR-EIA seems likely to replace the DCC method in the future.

Enzyme Immunoassay for PR 1719

REFERENCES 1. Allegra JC, Lippman ME, Thompson EB et al. Estrogen receptor status: an important

variable in predicting response to endocrine therapy in metastatic breast cancer. Eur J Cancer 1980, 16, 323-331.

2. Knight WA, Livingston RB, Gregory EJ, McGuire WL. Estrogen receptor as an indepen- dent prognostic factor for early recurrence in breast cancer. Cancer Res 1977,3’7,4669-4673.

3. Goussard J, Lechevrel C, Martin PM, Roussel G. Comparison of monoclonal antibodies and tritiated ligands for estrogen receptor assays in 241 breast cancer cytosols. Cancer Res 1986, 46, 4282-4287.

4. Leclercq G, Bojar H, Goussard J et al. Abbott monoclonal enzyme immunoassay measure- ment of estrogen receptors in human breast cancer: a European multicenter study. Cancer Res 1986, 46, 4233-4236.

5. Thorpe SM, Lykkesfeldt AE, Vinterby A, Lonsdorfer M. Quantitative immunological detection of estrogen receptors in nuclear pellets from human breast cancer biopsies. Cancer Res 1986, 46, 4251-4255.

6. Magdelenat H, Merle S, Zajdela A. Enzyme immunoassay of estrogen receptors in fine needle aspirates of breast tumors. Cancer Res 1986, 46, 4265-4267.

7. Greene GL, Harris K, Bovo R, Nolan C. Preparation and characterization of monoclonal antibodies to human progesterone receptor. Mol Endocrinol (in press).

8. Press MF, Greene GL. Localization of progesterone receptor with monoclonal antibodies to human progestin receptor. Endocrinology 1988, 122, 1165-l 175.

9. Scatchard G. The attractions of proteins for small ‘molecules and ions. Ann NY Acad Sci 1949, 51, 660-672.

10. Lowry OH, Rosebrough HJ, Farr AL, Randall RJ. Protein measurements with Fohn phenol reagent. J Biol Chem 1951, 193, 265-275.

11. Garola RE, McGuire WL. An improved assay for nuclear estrogen receptor in experimental and human breast cancer. Cancer Res 1977, 37, 3333-3337.

12. Bayard F, Damilano S, Robe1 P, Baulieu EE. Cytoplasmic and nuclear estradiol and progesterone receptors in human endometrium.] Clin Endocrinol Metab 1978, 46, 635-678.